Traditional antenna pattern measurement techniques utilize spherical scanning systems using mechanical or electronic positioning systems to move a transmitting and receiving antenna relative to each other in spherical coordinates. The positioning systems are generally disposed with two orthogonal axes of motion (e.g., theta and phi) to cover evenly spaced positions corresponding to the two angles of the spherical coordinate system. In any given implementation, each orthogonal positioner rotates either the antenna/device under test (AUT/DUT) in the center of the test volume, or the measurement antenna (MA) around the perimeter of the test range. The third spherical coordinate of radius (range length) is fixed for a given test implementation. The result is that the MA transcribes a sphere around the DUT in the DUT reference coordinates, with a radius defined as the range length, and always points toward the center of that sphere.
A third axis of polarization is also typically required along the radial direction between the center of the test volume and MA (i.e., along the axis of the MA). Polarization change is often performed electrically using a dual polarized antenna and switch, although a mechanical rotator can be used to rotate a single antenna element by ninety degrees about the axis of propagation from the center of the test volume. Likewise, one or both axes of spherical motion may be accomplished electrically by using measurement antennas in fixed locations and switching between them. In some implementations, two co-axial positioning approaches (i.e., a switched array and a mechanical positioner) are used to reach angles not accessible by a single solution (e.g., just the array).
These antenna pattern measurement techniques have been adapted for use in over-the-air (OTA) performance testing of wireless devices, whereby active radio frequency (RF) communication is carried out between the device under test (DUT) and a wireless communication tester connected to the MA. These tests are generally performed in a fully anechoic chamber to eliminate reflections and outside interference so that the pattern measurement test represents an average line-of-sight (LOS) performance of the DUT. Tests cover edge-of-link performance metrics of total radiated power (TRP) and total isotropic sensitivity (TIS) to determine the average transmitter and receiver performance in an over-the-air line-of-sight configuration. Note that the term RF as used herein means any frequency used for communication between radios, including microwave and millimeter wave frequencies.
Since most modern wireless devices utilize multiple antennas to address problems caused by multipath propagation in a real world environment, and even adapt to and benefit from such environments with concepts like multiple-input/multiple-output (MIMO) communication, new test systems have been developed for evaluating the performance of a DUT in an emulated multipath environment. An array of antennas distributed in two or three dimensions on the surface of a sphere about the DUT are fed simultaneously with the output of a spatial channel emulator to produce multi-cluster multipath scenarios, while the device is placed in different orientations representing typical usage cases within that environment. The result is no longer a radiation pattern of the DUT but rather an indication of the average performance of the DUT in the emulated environment. Tests have also moved from just evaluating edge-of-link platform sensitivity behavior to determining higher level multiple antenna performance in the presence of an unwanted interferer injected into the environment alongside the intended communication signals.
The existing over-the-air techniques were developed for existing wireless technologies; first for single input/single output (SISO) systems where all that was required was to use existing antenna pattern measurement (APM) techniques coupled with active communication testing, and later for MIMO and other multiple-antenna designs where multipath and spatial behaviors became an important part of overall radio performance. However, emerging wireless radio technologies are pushing the limits of even these systems and methods. Concepts like active beamforming and other adaptive antenna system approaches require that the test system be able to evaluate the combined performance of the radio, the antennas, and the software algorithms behind them. The technology is also moving towards a future where the behavior of the radio transceiver cannot be isolated from the antennas used and must therefore be tested completely in an over-the-air configuration.
The limitations in traditional SISO APM approaches are well known and require that the device under test be configured to generate a static antenna pattern in order to allow the measurement process to capture the radiated performance in each direction that represents a single snapshot in time of the device performance. Any adaptation that the device would otherwise perform as a function of the test process itself would invalidate the test results, giving an answer that had no meaning in terms of the real world performance of the device. This is because such adaptation violates fundamental assumptions of the antenna pattern measurement process, such as the fact that the two orthogonal polarization components measured at any given point represent components of the same field vector. If the device instead adapts to the change in polarization, the sum of those two components has no physical meaning, but rather is an artifact of the interaction between the test process and the DUT adaptation algorithm.
The boundary array RF environment emulation overcomes some of these problems by emulating a typical environment to which the DUT would be expected to adapt and evaluating the resulting performance. However, the boundary array also suffers practical limitations in both cost and complexity that impose limits on what is possible in current implementations. Primarily, the emulated environment is subject to Nyquist rules that increase the required number of antennas in the boundary array as the device size increases relative to a wavelength.
Proposed fifth generation (5G) mobile network technologies will rely heavily on beamforming techniques, both in the current bands where massive MIMO is intended to allow many simultaneous users in the same physical and spectrum space, and at the higher millimeter wave frequencies where current test technologies and propagation models don't necessarily apply.
For massive MIMO, large arrays with tens to hundreds of antennas will be needed to create simultaneous independent communication channels to different users. The physical size of these arrays will make typical boundary array test approaches difficult, and the beam forming adaptation that is the whole point of the technology cannot be tested properly in traditional APM based OTA test systems.
To overcome the path loss limitations that increase as a function of frequency, millimeter (mm) wave technologies are expected to rely heavily on beamforming techniques to find the best single path communication while also minimizing interference in order to increase the signal-to-noise-ratio (SNR) at the receiver. In addition to the narrow beam arrays that will exceed the Nyquist limits of any practical boundary array, the high frequencies of operation and broad bandwidths place limitations on the possible test equipment. At these frequencies, even a few feet of RF cable has too much loss to be viable.
To address these limitations, the industry needs equipment and measurement techniques that will allow the evaluation of active beamforming for both the desired signal path and any relative interferer at RF frequencies that include millimeter wave frequencies.